In our previous study we have demonstrated the direct relationship between the level of free and bound testosterone in plasma and HDLc. The association persisted after adjustment for age, cigarette smoking, alcohol consumption and obesity. Physical activity did not appear to be related to HDLc in this population because of the relatively sedentary status. We propose to re-evaluate the same participants two years later and again determine the relationship between hormones, the risk factors, and lipoprotein cholesterol. Measurement of both testosterone and estrogen will be done. A replication of our previous findings are necessary prior to doing more detailed lipoprotein and hormone studies. The men participating in this study are part of the Multiple Risk Factor Intervention Trial (MRFIT) cohort, and will be re-examined at their sixth annual examination. The cost of the study will be modest since most of the information is collected as part of the MRFIT study.